Sentri acts on brain’s impulses

The US Marine Corps has developed a robotic truck that mimics the brain’s ability to recognise different sounds.


The US Marine Corps has developed a robotic truck that mimics the brain’s ability to recognise different sounds, allowing it to identify gunshots and return fire.



The Gunslinger vehicle is based on the algorithm Sentri (Smart Sensor-Enabled Neural Threat Recognition and Identification), which mimics the electrical signals used by the brain to process sounds to recognise and locate the source of the noise.



Researchers at the University of Southern California’s Centre for Neural Engineering originally developed a map of the electrical currents in the hippocampus area of the brain, with the aim of using it to understand what people were saying in environments with a lot of background noise.



That proved too difficult, however, so technology company Safety Dynamics used the mapping diagram to identify the most simple of sounds: the gunshot.


Scientific programme officer Joel Davis from the Office of Naval Research, which funded the original researchers, said: ‘Most of them [researchers] were trying to recognise words in a noisy environment. It works in some situations and not in others. What’s really easy to distinguish is a gunshot.’



The system is being used by Chicago and Los AngelesCounty police departments in the form of cameras on lamp-posts and street signs. Both forces use wireless connections between the fixed camera positions and either local police stations or squad cars. When the Sentri system identifies a gunshot, its cameras zoom in on the position from which the gun was fired. They record the scene of the shooting and relay the footage to the police in real time over a secure internet connection.


Having adapted the technology for use in a robotic vehicle, the Marine Corps is making further enhancements to Gunslinger before deciding where the system would best be deployed.